Hard drive reliability study names names

In November, online backup provider Backblaze published some interesting statistics on hard drive mortality based on over 25,000 units in active service. It found that failure rates were higher in the first 18 months and after three years. Those conclusions matched the findings of other studies on the subject, but frustratingly, they didn’t include information on specific makes and models.

Today, Backblaze is naming names.

The firm has posted details on failure rates for 15 different consumer-grade hard drives, and the numbers don’t look good for Seagate. See for yourself:

Source: Backblaze

And that doesn’t even tell the whole story. In Backblaze’s storage pods, Seagate’s Barracuda 1.5TB has an annual failure rate of over 25%. The 5,400-RPM version of that drive fares better—its failure rate is only 10%—but that’s still pretty high compared to the competition. The failure rate of similar Hitachi drives in the same environment is less than 2%.

Only 10% of the hard drives in Backblaze’s storage pods come from WD, and they’re strictly low-power Green and Red models. The annual failure rates are pretty low, though: only 3-4%. Backblaze’s purchasing decisions are largely driven by price, which is probably why fewer WD drives are in the mix. They tend to be a little pricier.

Interestingly, two drives proved to be so unreliable in Backblaze’s storage pods that they were left out of the totals completely. Seagate’s Barracuda LP 2TB and WD’s Green 3TB "start accumulating errors as soon as they are put into production," the company says. It thinks vibration might be part of the problem. Other Barracuda LP and Green models seem unfazed, though.

Here’s a look at survival rates over time:

Source: Backblaze

After three years, only about three quarters of the Seagate drives remain. A surprising number of those failures come between 18 and 24 months, which contradicts the overall trend noted in Backblaze’s initial study. Infant mortality seems to be a bigger problem for the WD drives, while the Hitachis fail at a steady but slow rate.

Backblaze says the Seagate drives are also more prone to dropping out of RAID arrays prematurely. The company uses consumer-grade drives that aren’t designed explicitly for RAID environments, of course, but that doesn’t seem to bother the Hitachis. They spend just 0.01% of their time in so-called "trouble" states, compared to 0.17% for the WD drives and 0.28% for the Seagates.

Overall, Backblaze’s data suggests that Seagate drives are less reliable than their peers. That matches my own experiences with a much smaller sample size, and it may influence our future recommendations in the System Guide. Hmm. In the meantime, kudos to Backblaze for not only collecting this data, but also publishing a detailed breakdown.

Comments closed
    • mikato
    • 9 years ago

    HP does it in their servers also

    • mikato
    • 9 years ago

    I wish that hard disks were mentioned in that article.

    • jihadjoe
    • 9 years ago

    As a single case, no. But enough anecdotes and you basically have a survey.

    Not that I want to be the one compiling all that data, though.

    • Jason181
    • 9 years ago

    When flash cells fail as a result of endurance, they actually are no longer writeable, but are perfectly readable.

    • BIF
    • 9 years ago

    Ugh! The term “teachable moment” is like fingernails on a chalkboard. Even if it really IS a teachable moment, saying so makes you sound condescending.

    • Welch
    • 9 years ago

    I’ve had issue with drives from all brands. Specific drives from those brands also stick out in my mind, most of all being a Seagate Barracuda 2TB drive (Don’t recall if they were LP). However I think it best to take a step back and not jump on the kudos wagon to Backblaze. It is fair to say thank you to them for providing the data and not take it for granted but instead to take from it what is relevant.

    Things to keep in mind with the statistics provided by Backblaze.

    1.) This is specific to their environment and its hardware enclosures. – Its a possibility that their systems favor one drive over another. To completely ignore this possibility wouldn’t be wise, such as drives dropping out of RAID. It would be interesting to know which drives with which controllers.

    2.) The WD drives being used are Green and Red, which are much better suited for longer term storage than say a WD Black drive which is more targeted towards performance.

    3.) The number of WD drives used was only 10% of their total, so the testing methods aren’t really up to statistical testing par with the number of Seagate drives, that also means the Hitachi drives statistical data is near negligible in comparison.

    4.) Drives not being used in an environment that they were designed for. If the Seagate drives and some of the WD drives are models not designed to handle vibration or other RAID specific afflictions, then its close to a moot point in the first place. You can’t get upset if your Ford F-150 doesn’t get the gas mileage of a Toyota Prius, or that the Prius can’t tow like an F-150. There are specific uses for different drives, even if its only a firmware flash.

    5.) What firmware were these drives using… Its not uncommon for drives to have a firmware released to address reliability issues after release. While more common on SSD these days, HDDs still do receive some important updates.

    Take their article with a grain of salt. The entire finding was based on what is cheaper for RAID storage, nothing else. Performance wasn’t taken into account, warranty, single drive reliability or cost of replacement units if it was or wasn’t covered by warranty.

    Long story short, I feel that anyone basing their next HDD brand purchase off of this article alone without taking into accounts the numerous variables that accompany it, are doing themselves a serious dis-service. However, if your running an environment near identical to Backblaze, or your priorities are similar, they have provided you a template for running large scale data storage on the cheap. If your a home user or even I.T. consultant for small businesses, there is better criteria for choosing your next HDD purchase.

    • Diplomacy42
    • 9 years ago

    Tea party has good points? I think not.

    • davidbowser
    • 9 years ago

    Here is the report from 2007

    [url<]http://research.google.com/archive/disk_failures.pdf[/url<] They included PATA and SATA drives, but the test results ONLY has consumer grade drives. They don't give a reason that I could find in this document, but to flip-mode's comment, I have also read that Google ran tests that resulted in consumer drives having better value (performance/$/year). One of the interesting things that came out of the Google study was that SMART errors are insufficient indicators for predictive drive failure, as many drive failures had no SMART errors prior to failure. I wonder about that when looking at the SSD longevity test that TR is doing.

    • Visigoth
    • 9 years ago

    [url<]https://techreport.com/review/25889/the-ssd-endurance-experiment-500tb-update[/url<]

    • Visigoth
    • 9 years ago

    It explains how “albundy” just got 0wned! 😉

    • yeeeeman
    • 9 years ago

    Speaking of reliability, my 3.5inch 500GB Seagate 7200.12 is in agony after 5 years of use as a OS and data drive. Fortunately, it still runs and I can backup my data, but slowly it’s getting harder and harder to access it.
    On the other side, I have a Hitachi 7k200 on my notebook, running happily for the same amount of years. Giving the fact that it is in a much shaky environment, because of me carrying it around everywhere I need it, it’s still in very good condition. It has some reallocated sectors, but nothing to worry about.
    I also had expiriences with WD and they are not as bad as Seagate, but still, they are less reliable than Hitachi.

    • DarkMikaru
    • 9 years ago

    Yeah, I hear ya Skyline. But in the end, just like everything else it comes down to personal experience. A good friend of mine and I used to have this debate all the time back in the day. To me, it seemed like every WD drive I touched would fail while I had several Maxtor / Seagate drives live on for years. Oddly, for him, it was the exact opposite. You couldn’t get him to touch a Maxtor / Seagate drive if his life depended on it.

    I have several dead drives that need to get recycled on my desk. When I get home from work I’ll take a picture, count them up and come back and post the results. I’m fairly certain they are mostly dead seagate drives along with WD & Fujitsu also.

    With regards to reliability? I had my suspicions that the “enterprise” drives were, in reality, no better than any consumer version you could buy off the shelf. Short of a longer warranty… they are the same.

    Also, been meaning to ask… your name isn’t based off a Nissan Skyline R33 GTR is it? Anytime I see you post thats exactly what I think. So therefore I’ll just name you Skyline. 🙂 The gerbil has spoken.

    • LaChupacabra
    • 9 years ago

    If you could cite a source for that it’d go a long way toward your credibility. The way that bits get written to flash in those is like a giant raid array. Hard disk recovery is a known and proven (although) expensive thing. Recovering data off an SSD that has had a flash failure (and correct me if I’m wrong here) is impossible.

    • just brew it!
    • 9 years ago

    A lot of rackmount servers (including our Supermicro here at work) mount the drives on their sides as well. We also have a few HP desktops that mount the drives vertically (standing on end). So major OEMs believe that it is OK to mount drives in a non-flat orientation.

    • JosiahBradley
    • 9 years ago

    But when an SSD dies, recovery of the NAND should be significantly easier than examining an iron disk under a microscope. Or just swapping out controllers, which is more of a pain with a HDD because the daughter-board normally is attached to the power delivery system.

    • dlenmn
    • 9 years ago

    That’s a good point, but while most PCs mount HDs flat, it’s not _that_ unusual to mount HDs vertically or on their side (all-in-ones, my Silverstone case, etc.). It would be interesting to do a flat vs non-flat mounting test.

    • Krogoth
    • 9 years ago

    You realize that SSDs have their own set of issues related to longevity?

    • JosiahBradley
    • 9 years ago

    That’s it, I’m not going to buy anymore hard drives. Solid state for me from now on. I’m in that small percent of people who don’t need too many terabytes. So when terabyte ssds come down in price, I’ll be getting those exclusively.

    • egon
    • 9 years ago

    Interesting, but over time I’ve become less concerned about drives dying outright and more curious about rates of ‘bitrot’ and silent data corruption in general, because nowadays I’m already prepared for outright failures and other obvious threats (e.g. fire, burglary, power surge) with a couple of extra drives and onsite/offsite backup routines, but am left with zero defence against silent corruption.

    • lhl
    • 9 years ago

    Actually, I think you’re the one that’s being dense. Completely, missing the point, certainly. And getting MAD (ALL CAPS), as well it seems. I’m not the biggest believer in arguing on the Internet, but maybe this is a teachable moment. Backblaze’s results have nothing to do with features, or operating conditions, or anything that you’ve mentioned in this whole thread. It’s just Statistics 101.

    The numbers of drives that Backblaze is running is far beyond what you need to get a good confidence interval for failure rates. I plugged in the smaller WD number (3973 drive years) and given their stats, their margin of error is +/0.5%. With the Seagate drives, this is closer to +/-0.3%

    On Hacker News one of the Backblaze guys mentioned that they literally plug their numbers (capital costs, operating costs, failure rates) into a spreadsheet and it tells them what to buy. And that’s exactly the right thing for them to do looking purely from the perspective of storage cost vs dead drives.

    They spend a whole paragraph of disclaimers saying that their numbers apply only for their specific environments, so I don’t know what you’re getting at. Maybe if more people released/pooled their drive attrition rate in various operating environments we’d get a better picture of how various drives fare in various conditions, but no one else has published jack. Backblaze has, so I say kudos to them.

    So by all means, please publish your own results/blog post, but keep in mind that the plural of anecdote is not data – well, unless it’s a valid set of statistically significant number of anecdotes. Then it very much is data.

    • Deanjo
    • 9 years ago

    I don’t know, I have more Maxtors still kicking than any Seagates. 1st gen SATA’s going hard on a daily build server 24/7 and still going strong.

    • Bensam123
    • 9 years ago

    Maxtors destructive powers linger on!

    • Bensam123
    • 9 years ago

    Yeah, I’ll probably start buying Hitachis now… I never used to even look at them. It was either Seagate or WD.

    • hbarnwheeler
    • 9 years ago

    I don’t see them blaming a manufacturer, or penalizing an entire line, though I don’t quite know what you mean in asserting that this is what they’re doing.

    Granted, these are consumer drives not intended to be used in an environment with this much heat or vibration, but heat and vibration tend to be what kills hard drives, so it is useful data for me, a consumer, when making a purchasing decision. Here comes the car analogy: If one manufacturer’s cars fall apart when used in an environment harsher than that for which they were not intended, and another manufacturer’s do not, it seems perfectly rational to think that a car from the second manufacturer will fare better over time (though, it would be nice if the WD sample were larger).

    It may be that Seagate’s newer (or higher priced) drives would fair better, but it doesn’t follow from this possibility that it would be irrational for a consumer to take this Backblaze data into consideration . Barring knowledge of what has changed in the newer drives and the likelihood that these changes will increase reliability, this data seems to be among the best available, though by no means perfect, as you and other have pointed out.

    • Meadows
    • 9 years ago

    I’m amused by how wrong you can get sometimes.

    • Meadows
    • 9 years ago

    I blame today’s videogames.

    • cynan
    • 9 years ago

    Don’t you mean rate? Wouldn’t ratio imply there were instances where the infant did the killing?

    • squeeb
    • 9 years ago

    The old Samsung spinpoint series was baller. Too bad they sold off their mechanical division. I still have 2x 2TB F3s in my NAS running 24/7 for almost 3 years now.

    • Jason181
    • 9 years ago

    If margins are razor thin, then cutting corners is not so much keeping profit margins as high as possible, as [i<]having[/i<] a profit margin.

    • cynan
    • 9 years ago

    It’s not unfair, but it may represent a distorted picture of all Seagate drives, and therefore be of questionable value to a consumer purchasing HDDs now.

    As I’ve [url=https://techreport.com/news/25940/hard-drive-reliability-study-names-names?post=794843<]written elsewhere[/url<], the bulk of these 1.5 and 3TB Seagates seem to be specific models plagued by firmware issues.

    • jss21382
    • 9 years ago

    I have 3 dead 1.5tb seagates I bought during my last upgrade cycle, and a couple dead green drives from the cycle before that, so these results seem pretty valid to me.

    It’s hard to say consumers buy the cheapest crap possible in the hard drive space. It’s not like we have value hard drive makers, Seagate and WD respectively were both great brands at the time.

    With SSDs there’s a heirarchy between the brands and models, In HDs it’s been years since there was a meaningful difference from one brand to another, and from one line to another aside from the Raptors.

    • Milo Burke
    • 9 years ago

    Don’t consumers always buy the cheapest crap possible anyway?

    And why should we have to buy mid or high range mechanical drives for more reliability when we can just purchase select models or brands in the budget range that already have significantly higher reliability than others?

    • playboysmoov
    • 9 years ago

    Actually, Seagate purchased Samsung’s disk based drives and Samsung is still making their SSD.

    During my experiences I have had drives from WD, Seagate and Samsung fail. I currently have a WD Caviar Green drive 1 TB drive I moving data off now because of Smart Errors in WDLG tools.
    YMMV.

    • Waco
    • 9 years ago

    IMO, you’re spouting a lot of opinions with little to no basis in fact or experience.

    As someone who manages many thousands of HDDs in RAID arrays on a daily basis, consumer drives fail just as often as enterprise drives (even in non-vibration proof JBOD enclosures). Their failure modes tend to be different but overall they fail at nearly the same rate as expensive enterprise-level disks.

    Sure, their performance is more sporadic, but calling BackBlaze a bunch of incompetent idiots is both ignorant and incorrect.

    • bwcbiz
    • 9 years ago

    I’d like to see a reliability comparison of Seagate vs. Samsung. Ever since Samsung bought Seagate and they started sharing model numbers, I’ve been getting the impression that Samsung is cherry-picking the best hard drives for release under their own brand name and leaving the ones that barely pass QA to the Seagate brand. The Seagate numbers here are suggestive, but they only tell half the story if that’s what’s going on.

    • indeego
    • 9 years ago

    “Why do we have the drives we have? Basically, we buy the least expensive drives that will work.”

    So they pick a sample of very few drives and then blame the manufacturer, rather than the model?

    I wonder what models fail at the mid or high range?

    This isn’t a reliability study inasmuch a “we bought the cheapest crap possible and guess what, one manufacturer’s cheap line of drives failed more than the others.”

    What a surprise!

    But to penalize the entire line is silly. It’s like criticizing 2014 imacs because you had a bad ipod a decade ago.

    • maxxcool
    • 9 years ago

    😉 car ran great till it didn’t!

    • maxxcool
    • 9 years ago

    Ah!! big words! rendered … helpless… Loiissssss heeelllp meee!

    • just brew it!
    • 9 years ago

    Sounds like Seagate already burned them by shipping a bunch of dodgy refurbs as warranty replacements. They may actually be doing this to try to get Seagate’s attention and gain some leverage.

    • Milo Burke
    • 9 years ago

    I would love to see a source for this. Anyone know one?

    • Veerappan
    • 9 years ago

    This. I’ve got a 1-2GB USB stick plugged into the back of the machine as its boot drive and it’s working fine. I spent all of my budget on redundant storage of the data that’s actually at stake.

    • Bauxite
    • 9 years ago

    The hitachi plant that got to keep the name is in china.

    • Bauxite
    • 9 years ago

    Before samsung was [s<]gobbled up by the duopoly[/s<] went away, anecdotally they were on my list with hitachi as the drives to buy. (~dozen 2TB in my ZFS box at the moment) What this really proves is the people that forced WD to give up some of the hitachi plants in the buyout were very wise to the real effects of market shrinkage.

    • circus108
    • 9 years ago

    Perhaps the result of this will elevate competition above the religiosity/trance of “branded names”, thereby driving “Seagate” to up the warranty/confidence-index ante. The most to fear from this is that quiet mouse in the corner reaping from the harvesting of, “Ve von’t take NO fur zer answer”, You know, that propagandist par excellence and its “Intellipower” mumbo-jumbo. Perhaps it’s time to let loose a little Intelligence into the HDD arena thereby knocking Religion & Religiosity from the HDD Totem Pole.

    • stdRaichu
    • 9 years ago

    Well I still found it odd that they run the risk of pissing off Seagate whilst still using their drives as their primary new storage devices as they state they do with the 4TB ones 🙂

    No idea on wdidle3 with the latest greens I’m afraid – I assume you mean the SATA 6Gbps/FZEX models? I’ve not bought a green for a while since the reds came out. TLER was too good to miss out on for anything in RAID…! At a pinch I’d say check your LCC counter and other SMART attrs if you choose to install one but get a hitachi or a red if you’re unsure.

    Edit: I’ve found out there’s an open source alternative that sets the same things as wdidle3 without the need to boot into a DOS environment (although the discs do need to be powered down and back up again for the setting to take effect). It’s available as idle3-tools in debian.
    [url<]http://idle3-tools.sourceforge.net/[/url<]

    • jihadjoe
    • 9 years ago

    Infant Kill Ratio?

    • Mazo
    • 9 years ago

    I believe that was sold off to Western Digital in 2012.

    • l33t-g4m3r
    • 9 years ago

    Are you seriously that dense? I answered that in my first post.

    1. Only 10% of the hard drives in Backblaze’s storage pods come from WD
    2. Backblaze’s purchasing decisions are largely driven by price
    3. The company uses consumer-grade drives that aren’t designed explicitly for RAID environments

    They’re PRIMARILY using consumer Seagate hard drives in a RAID environment, so the numbers are MASSIVELY SKEWED, and they admit their “PODs” are not ideal operating conditions. All this data proves is that consumer Seagate drives don’t work well in Backblaze’s system. It doesn’t say anything else, because they didn’t use an equal amount of competitive drives, nor was it under proper operating conditions.
    [quote<]We think this is related to vibration. The drives do somewhat better in the new low-vibration Backblaze Storage Pod[/quote<] Basically, Backblaze is only naming names because they wanted to be cheap and use desktop drives in an enterprise environment, and they think doing so will get Seagate to put enterprise features in consumer drives for free, because of the publicity. At the end of the article, Backblaze says it's going to CONTINUE buying consumer Seagate drives, so there's obviously a hidden motive for publishing the article in this manner. They also say their new WD drives will be the RED editions, which are NAS drives with anti-vibration features. Funny how they aren't buying similar Seagate drives, and WD's red drives are the ones with poor ratings on newegg. (look it up.) From WD: [quote<]3D Active Balance Plus The enhanced dual-plane balance control technology significantly improves the overall drive performance and reliability. Hard drives that are not properly balanced may cause excessive vibration and noise in a multi-drive system, reduce the hard drive life span, and degrade the performance over time. [/quote<] So is WD WRONG then? I think not. These features are necessary in a RAID environment. Either way, there is apparently some defects in the RED's according to the newegg reviews, and if that is still the case Backblaze's next "expose" will be on the WD RED's. IMO, they're a bunch of incompetent idiots using the wrong drives for the wrong situation.

    • lhl
    • 9 years ago

    What’s unreliable about their data? Regardless of whatever features enterprise drives might or might not have vs consumer drives, if their failure rates are the same across a large enough sample size, that’s that’s the way it is, regardless of anyone’s opinions. On the other hand, you’ve seemed to offer lots of opinions, but without any data. Well I guess it *is* the Internet…

    • psuedonymous
    • 9 years ago

    One thing that [i<]might[/i<] be a factor is Backblaze's 'storage pods' all orient the drives vertically, interface-down. This means that the platter bearings are worn on an axis perpendicular to use in the normal horizontal orientation (if you're using the wrong bearings, this could result in diminished lifespan), and the seek-arm is pointing directly upward rather than balanced on it's pivot (may be more vulnerable to seek errors without gravity damping head-bounce if the pivot bearing has pitch/roll slop). This in no way invalidates the results (both these factors are only an effect if corners are cut compared to other drives), but it may mean that Seagate's high drive failure rate may be less pronounced with horizontally oriented drives.

    • Pettytheft
    • 9 years ago

    Maxtor.

    • just brew it!
    • 9 years ago

    The problematic drives in that study appear to be from the 7200.11 generation. “What happened” is that they got unlucky and used a boatload of Seagate’s worst drives.

    I do prefer HGST/Toshiba and WD over Seagate these days, but not enough that price can’t sway my purchase decision. IMO Seagate is in the same situation today that IBM/HGST was in immediately post-Deathstar: QA issues are generally under control, but prices are artificially depressed relative to the competition due to the lingering hit to their reputation. IOW, good bang for the buck!

    • just brew it!
    • 9 years ago

    Hitachi’s HDD division was acquired from IBM (an American company), and the drives are manufactured in China and Thailand. So what does that explain, exactly?

    • just brew it!
    • 9 years ago

    While server rooms do tend to be less variable in terms of ambient air temp, and server racks tend not to get banged around while the drives are running, you do tend to have a lot of equipment packed closely together, and vibration from nearby drives in a RAID array can be problematic. So I wouldn’t be so sure that the environment experienced by individual drives is *that* much better.

    • just brew it!
    • 9 years ago

    FWIW Seagate has about 6x the market capitalization of AMD.

    • just brew it!
    • 9 years ago

    For a while (for a few years after the “Deathstar” mess) they were substantially cheaper. They were an incredible bang for the buck back then.

    • just brew it!
    • 9 years ago

    The Seagate drives in that study are all ~4 years old, either 7200.11s or of similar vintage. I believe they have cleaned up their act quite a bit since then. My 4 7200.12s have been fine (two of them have been in use almost nonstop for 3 years).

    • just brew it!
    • 9 years ago

    Unless the drive dies early on, the value of the warranty is dwarfed by the inconvenience of swapping the drive out, dealing with the RMA, and recovering the data from a backup (assuming you have one).

    • just brew it!
    • 9 years ago

    Yup. Most (all?) of Toshiba’s current 3.5″ desktop drives are re-badged Hitachi drives. The 2 that report as Toshiba are probably still Hitachi designs, they just finally got around to changing the drive ID in the firmware.

    I work for a business unit that got bought out a year ago. Some of our documentation and firmware still claims to be from the old parent company. It takes a long time to change all of that sh*t over, and it gets lower priority than actually developing/shipping working product that you’re getting paid for!

    • just brew it!
    • 9 years ago

    Nope. China and Thailand mostly (or entirely, not sure… basing this on drives I personally own).

    • just brew it!
    • 9 years ago

    What generation? There were only a couple that were problematic. I still have numerous drives bearing the “Deskstar” brand name that are working fine; they are all from after the problematic “Deathstar” models.

    • just brew it!
    • 9 years ago

    A quick spot check of the labels on several Hitachi drives I have laying around indicates that they have factories in China and Thailand. There may be others as well.

    Interestingly, the *only* Hitachi drive I’ve had trouble with in the past decade was one of their laptop drives. Don’t know where it was manufactured, it is long gone.

    • just brew it!
    • 9 years ago

    Every vendor has their problematic product generations. IBM (who later sold their HDD business to HGST) had their “Deathstar”. Maxtor had the DiamondMax 9 (and at least one other problematic generation way back in the day). Seagate had the multiple 7200.11 fiascos. WD had a turkey or two as well, I just don’t recall the specific product families.

    The pattern seems to be: HDD maker gets complacent, cuts corners, and puts out a generation of drives that suck. Then they start paying more attention to quality, and things get better for a while. Post-“Deathstar” IBM/HGST drives have been good. Post-7200.11 Seagates seem pretty reasonable as well in my experience. WD has been reasonably good these past few years as well, other than the head park timer issue on the Greens.

    • just brew it!
    • 9 years ago

    Yes, but HGST’s 3.5″ business got spun off to Toshiba after WD acquired it, so it ain’t that simple!

    • just brew it!
    • 9 years ago

    Or just don’t care. As long as they are reporting factual data Seagate would have a tough time prevailing in a lawsuit.

    • just brew it!
    • 9 years ago

    The drives they “shamed and named” are from several product generations prior to yours (7200.11 era). My 7200.12s have been OK. Your drives are even a gen or two newer than that.

    • just brew it!
    • 9 years ago

    Sure. But consumer desktops can be subjected to heat and vibration as well. The typical living room, den, or office may not have as good of climate control as a server room, and computers/drives located on/under desks are susceptible to being bumped, kicked, knocked over, etc. so there’s definitely the potential for mechanical shock while they are operating.

    • just brew it!
    • 9 years ago

    Not as ironic as you might think, and only half right.

    After the “Deathstar” debacle they were paying a lot more attention to QA (and with IBM’s deep pockets they could afford to do so). They were also pricing their drives very aggressively to win back market share. So it does not surprise me that they were very reliable (and great bang-for-the-buck on top of that).

    Hitachi does not own the tech any more though. They sold their HDD business to WD, who was in turn forced to sell off the 3.5″ desktop drive portion to Toshiba to satisfy anti-trust regulators in various countries.

    I think I will be buying more Toshiba drives going forward…

    • just brew it!
    • 9 years ago

    [quote<]I'm curious as to what prompted the change in policy from "no names" to "names and shames".[/quote<] Probably not so much a change in policy, as someone who doesn't care whether they piss off Seagate. Agree with you on the Hitachi drives; I've had very good luck with them these past ~10 years. Incidentally, do you know if the current crop of WD Greens still need/support use of wdidle3?

    • ozzuneoj
    • 9 years ago

    On a somewhat related note… has anyone here noticed a trend with regards to the reliability of different brands of 2.5″ drives in laptops? I’ve seen several Toshiba drives from the Vista and Windows 7 days with terrible problems, where as most of the others I see that are problem free are Hitachi.

    Most systems from that period (2006-2010) were prone to heat problems if not used carefully, which certainly wouldn’t help matters, but I have seen a slight trend of Hitachi being more reliable in this situation as well.

    • willyolio
    • 9 years ago

    FreeNAS OS… drive?

    y u no usb memory stick?

    • willyolio
    • 9 years ago

    it’s still nearly 3000 WD drives. that’s more than sufficient sample size to give an overall reliability number for the manufacturer. as long as they weren’t all purchased in a single batch.

    • draconian
    • 9 years ago

    How does this compare to failure rates for SSDs?

    • l33t-g4m3r
    • 9 years ago

    I’ve also had several WD’s catastrophically fail on me, right around or before a year. Seagate hasn’t given me anywhere near such trouble, as I only have had one 1.5 fail, but it gave me audible warning signs and I was able to save the data. I still have some of the original 1.5’s going that needed the firmware update, and I’ve had no issues whatsoever with their newer 2gb drive.

    When reliability is primarily based on anecdotal evidence, I buy what is the best value, which is usually Seagate. WD charges too much, nor have they been any more reliable in my personal use. Maxtor was the same way, I never had one of those drives fail, and the maxlines were perfect for desktop use. IMO, their problem was that too many distributors were selling refurbished drives as new.

    • hbarnwheeler
    • 9 years ago

    How is publishing these results unfair to Seagate? It is not as though they dropped the Seagates off a roof first. They just bought thousands of new drives, used them, and reported which ones failed, and when.

    • Krogoth
    • 9 years ago

    Blame the bean counters and upper management finding ways to keep profit margins as high as possible in HDD market where such margins are already razor thin. At Seagate, they figured that they can make more than what would they lose to warranty work and lost of customers from failed units. It happens more often than you think which is why I put in current Toyota and glory-days era GM as notable examples.

    It is mostly Seagate’s customer-tier line. Their enterprise-tier stuff appears to be far more respectful. WD doesn’t fare that much better in the customer-tier arena. Hitachi units in the sampling were enterprise-tier so that made them look a little better where most of the Seagate samples were customer-tier. That’s why Seagate appears to look worse. They also had larger populations in the samplings as well.

    FYI, I operate and own a number of Seagate HDDs. I only had one unit failed in its lifetime. I know people that had several failed WD units. I had one WD Raptor which was dying that I had to get RMA for a refurb unit.

    • DarkMikaru
    • 9 years ago

    Sorry to change the subject… but man, those were good times. $40 for a 750GB drive? I remember buying 2 WD Green 750’s off Newegg for 39.99 free shipping. Those bad boys are still cranking away. 🙂

    But yeah, I agree. Seagate used to be my default as WD drives always seemed to die on me. Now, its Seagates that I cant get to live if my life depended on it.

    • l33t-g4m3r
    • 9 years ago

    Consumer drives don’t support TLER, or have anti-vibration capabilities. Do you really [i<]need[/i<] those features? Maybe not, but they certainly are designed to increase longevity in a RAID environment. Considering that Backblaze says the only difference between the two is the warranty, I don't really think these guys are a reliable source of information. Their data is skewed from their purchases, and so is their opinions.

    • FuturePastNow
    • 9 years ago

    I liked ’em fine until they started failing.

    • Bensam123
    • 9 years ago

    I used to think so to, I bought four of them. One of them died this year, another is getting really flaky. I have used Seagate almost exclusively over the year, but my opinion on them has been changing, especially in the last few years.

    • cynan
    • 9 years ago

    It seems to be that many of the 1.5TB drives were from the 7200.11 erra, where a certain firmware (can’t remember which), led to these drives failing. Backblaze reported buying larger Seagates in recent years…

    I wonder how many of those 3TB Seagates (from the blog, most seem to be the 7200.14 ST3000dm001) had issues due to firmware? There is a [url=http://ask.adaptec.com/app/answers/detail/a_id/17241<]known issue[/url<] where earlier 7200.14 barracudas, with firmware CC4G or earlier reported errors or were unstable in RAID arrays. I have 3x the 2TB versions in a NAS that have been running for 1.5 yrs without issue - and I remember making sure to upgrade the firmware on all before I deployed them. And this makes me question how much of the issues with Seagates, pictured here, are actually manufacturing QC issues or due to bad firmware... It would be interesting to see these results again in a couple of years. Given the above, I think Seagate has a good chance of looking much better (though probably not quite up to Hitachi here).

    • Kougar
    • 9 years ago

    I bought a few junk drives to take apart for a project. One ended up being a failed Deathstar drive, and noticing it was still well under warranty I sent it in. The replacement drive still works fine today… can’t say the same for some newer drives I bought since then.

    • cynan
    • 9 years ago

    I’ve had 2 of those 640GB WD drives running in RAID 0 since 2008. Still going…

    • continuum
    • 9 years ago

    Would be very curious to see a longer-term report in the future on 4TB drive reliability. We’ve known Seagate’s 1.5TB drives were lemons long before this report…

    • Waco
    • 9 years ago

    It was luck of the draw on the firmware really – I had one of them die with the bad firmware but the other had the newer version that was “fixed”. It’s still cranking 24/7 in my file server without any apparent issues.

    Anecdotal, I know, but the Backblaze data seems to agree with it.

    • captaintrav
    • 9 years ago

    I think all of the 7200.11s, I had a 1TB one, after the second replacement died I gave up. 5 year warranty is pointless when it can’t even last one without taking a dump.

    • hasseb64
    • 9 years ago

    Good story! Maybe one of the best for a long time.
    Never bought any seagates (been building for 20 years), my stomach was right..

    • Bensam123
    • 9 years ago

    Ever since the 1.5TB Baracuda 7200.11s. You used to be a great company with top tier reliability. What happened?

    • adisor19
    • 9 years ago

    When they dropped their famous 5 year warranty, i switched immediately to WD Black.

    It was a sign that they were not confident in the life of their drives.

    Adi

    • GasBandit
    • 9 years ago

    I used to be a hard core Seagate supporter, went years of buying nothing but Seagates. But I did notice once their drives started hitting the 1+TB mark that quality and reliability went straight to hell. Even Western “Ping of Death” Digital has failed on me less in the last 5-7 years. Looks like though, by these charts, I’ll be taking a closer look at Hitachi in the future!

    • Palek
    • 9 years ago

    [quote<]The only striking detail that I remember about those drives was that they were made in Hungary.[/quote<] This is what I remember as well. IBM shut down the Hungarian factory right after the Deathstar debacle. They also made Deskstars in Singapore or Malaysia, and those drives were unaffected if I remember correctly.

    • lhl
    • 9 years ago

    Ah, you’re probably thinking of the old Google drive report: [url<]http://static.googleusercontent.com/media/research.google.com/en/us/archive/disk_failures.pdf[/url<] Google standardized on "consumer-grade disk drives" but I don't think they actually published their comparison vs enterprise drives, although the assumption was the difference must be negligible if they were running 100K's of consumer drives in all their data centers...

    • Pwnstar
    • 9 years ago

    Tea party has good points, though, so your post is a non sequitur.

    • Growler
    • 9 years ago

    Wow. This takes me back. I had a 13GB Deskstar in the first computer I built. I got it because of the review in [i<]boot[/i<] that raved about it's awesome ATA 33 performance. Combined with my 1.1 Ghz Athlon, I had a pretty nice box back then. Ahh, memories.

    • smilingcrow
    • 9 years ago

    “I have a 1TB WD green drive since 2008, pretty much on 24/7 as a DVR drive.
    Probably done enough writes to trash a Samsung EVO twice over”

    If you record video 24/7 at 24 Mbps for 6 years that equates to 540TB which is close to what a 250GB EVO has already managed in the TR endurance test; a 1TB version will endure 4 times longer.
    So a 1TB Evo recording at 24 Mbps for 8 hours a day will reach the equivalent write amount in ~72 years.

    • Airmantharp
    • 9 years ago

    As if any of these are fully manufactured in the U.S. or Japan.

    • Airmantharp
    • 9 years ago

    Make sure you put it in a room with automatic fire suppression 🙂

    • Waco
    • 9 years ago

    Anecdotal evidence isn’t really helpful. I have a pile of dead Seagate drives that “prove” the exact opposite if you want to trust anecdotes. 🙂

    • FireGryphon
    • 9 years ago

    I used a Deathstar drive for years, and only stopped using it because I wanted a drive with more capacity. It was working fine.

    • chrissodey
    • 9 years ago

    This pretty much mirrors what I have seen over the years with Lenovo laptops. During the T400 days we had to have about 40% of the hard drives replaced within the 3 year warranty period. About 95% of those drives were made by Seagate.

    In the early 2000’s I only bought Seagate because there was a time were all their drives had 5 year warranties. Sadly those days are gone. My favorite drives so far are the WD 640GB that were made just before they switched to a color branding scheme and the Samsung 1TB F1. Unfortunately Seagate killed the great Samsung F1 and F3 drives.

    • SoM
    • 9 years ago

    i meant currently only have one seagate in the PC, but had other sized seagates that also never failed, compared to hitachi/fujitsu/maxtor that gave me problems.

    my oldest 8.4gb seagate still works 🙂

    • Deijya
    • 9 years ago

    My 1tb western digital black that I bought back in ’08 is still in my upgraded system as a secondary drive to my ssd samsung 256 pro. It survived at least a dozen crashes, several hard boots caused by my cat hitting the reset button, and the weird win7 to win8 to win8.1 upgrade issue I had. I don’t do a lot of read/writes with it though, so that would help its longevity.

    • sschaem
    • 9 years ago

    I have a 1TB WD green drive since 2008, pretty much on 24/7 as a DVR drive.
    Probably done enough writes to trash a Samsung EVO twice over, and its still going strong.

    But I did have a 2TB WD drive fail (one out of 4) in its warranty period.

    Now I wonder what will happen first ? an HDD failure or 2TB MLC SSD dropping below $100?
    …Sadly my bet is on another HDD failure.

    • albundy
    • 9 years ago

    considering that Hitachi is Japanese, and the rest is American, it really explains the quality of the brand. it’s really too bad WD lowered their rep. when they bought them.

    • SomeOtherGeek
    • 9 years ago

    Yea, good idea. My rule of thumb is don’t trust anything computer. Anything can good, so just backup everything or have a backup ready.

    • NovusBogus
    • 9 years ago

    Shazbot! And here I was thinking Seagate’s reputation was just another case of Internet emorage, but it does appear they’ve earned it. Too bad Hitachi barely exists outside of the enterprise/NAS world anymore, I loved my 1TB deathstar but the current models have received a lot of complaints.

    • flip-mode
    • 9 years ago

    Sweet. Maybe that’s the one I was thinking of.

    • Deanjo
    • 9 years ago

    Actually I believe it was only the desktop drive division that was sold to Toshiba, the 2.5 division was retained by WD.

    • ozzuneoj
    • 9 years ago

    This is pretty interesting…

    Only 10% of the drives were WD (as mentioned), but Seagate and Hitachi are about equal at around 45%. That makes the results even more significant IMO, as the playing field is relatively even. Also, the drive with the 25% failure rate is a 7200.11 model. I remember these being notoriously unreliable, especially after Seagate’s breakthrough 7200.10 drives, which introduced perpendicular recording and were some of the best drives available in 2006:
    [url<]https://techreport.com/review/10063/seagate-barracuda-7200-10-hard-drive[/url<] They've certainly gone down hill since then though. I own several 7200.10s, and I pick up used ones often for fixing up older systems and they are rock solid. My 7200.12 500Gb had a ton of bad sectors only a year or so into its life though. My now 4-5 year old Samsung F1 1Tb has 100% perfect health and performance to this day. Its a shame all of the best mechanical drives are no longer produced. That said, I will definitely be looking at Hitachi much more closely now, and you can bet I'll be keeping a copy of their reliability chart for when I'm buying used drives.

    • jessterman21
    • 9 years ago

    IKR, I bought a 750GB Deskstar back during flood-prices because it was the cheapest, and I was really worried about the reliability. No more!

    • cynan
    • 9 years ago

    Sure, this looks bad, but this is essentially a cherry-picking of a few drive models.

    This doesn’t mean that Seagate is worse in general, today, than any other brand currently available – which is the info that is really important to customers.

    • DarkMikaru
    • 9 years ago

    I can believe it! I stopped buying Seagate drives for my customer & personal builds years ago. Why? Failure rates were crazy and it seems to corroborate what this & many other studies have observed along with my own observations. If the drives survive the first 3 to 6 months of use then they’ll likely go on to live long productive lives. The last build I took a chance on with a Seagate was 4 months ago. The drive died after 1 month exactly! The one before that died exactly 2 months in. The sad thing is it seems notebook & desktop drives have virtually the same expectancy. In my experience they both seem to die at about the same rate.

    I’ve since gone WD Greens for my home drives, WD blue for customer & friend builds and guess what. Not a single one of them has died. I have several 3 to 5yr old Green drives that are still running 24/7 in my home server. One last thing, I’ve also noticed Hitachi drives being DOA out of the box or living forever, kinda hit or miss.

    Great report TR..thanks for assuring me I am not crazy.

    • NeelyCam
    • 9 years ago

    I wanna make an 8-disk RAID0 array with deathstars and see how long it lasts

    • adisor19
    • 9 years ago

    Samsung had its ups and downs. Just like Fujitsu and Quantum before that.

    Adi

    • NeelyCam
    • 9 years ago

    He’s with the Tea Party.

    • indeego
    • 9 years ago

    > failure rates for 15 different consumer-grade hard drives,

    But there are HUNDREDS of drives out there. (Newegg shows 581 different types of internal drives for sale as I type this.) This is hardly fair to Seagate.

    • Deanjo
    • 9 years ago

    Muhahahaha, told everyone that Seagates were horrible for reliability and that they were by far the maker with the most failures at Apple but ohhhhhhh nooooooooo would anyone believe me?

    • hubick
    • 9 years ago

    I also had 6 of those drives spread across 3 separate RAID-1 arrays inside well-ventilated boxes, and none of mine failed before I upgraded to larger ones.

    • lhl
    • 9 years ago

    Actually, your intuition is wrong. Backblaze published an article last month comparing a statistically significant number of enterprise vs consumer drives and found no difference in failure rates. The only advantage that the enterprise drives had was warranty period: [url<]http://blog.backblaze.com/2013/12/04/enterprise-drive-reliability/[/url<] The analyses Backblaze post are awesome for exactly this reason - proving/disproving widely held assumptions and because no one else seems to publish numbers about how different brands/drives fare.

    • dmjifn
    • 9 years ago

    Well, they live on under Seagate… if that makes you feel any better. 🙂

    • dmjifn
    • 9 years ago

    I think they probably aren’t across the board. I bought 2 hitachis at Frys late summer and they had the same model on the shelf coming from both China and Thailand. One of them said Hitachi and the other said HGST, otherwise it was the same box.

    • csxcsx
    • 9 years ago

    If you go to the Backblaze link, they have all the model numbers.

    • flip-mode
    • 9 years ago

    Interesting. I have always suspected WD had the worst reliability.

    High-five to Backblaze for naming names. Too bad the old Google hard drive study did not do so.

    • Scrotos
    • 9 years ago

    Let’s not go crazy, here.

    • flip-mode
    • 9 years ago

    I forget where I read it, but I did read that someone (google, I think ???) uses consumer drives because “server-grade” drives are not any more reliable.

    • odizzido
    • 9 years ago

    Yeah my seagate drives have always died too. I did get a 500gig seagate drive for an OS drive a few years back because all the good HDs were all sold out and it has been a number of years since I have gotten a seagate drive. Surprisingly it still works.

    • Wildchild
    • 9 years ago

    I miss Samsung hard drives.

    • UnfriendlyFire
    • 9 years ago

    This calls for a SSD reliability study. I’m not talking about the wear-downs, TR already got that.

    I’m talking about other kind of failures. Controller issues, buggy firmware bricking the SSD, damaged NAND die that somehow passed/avoided inspection, bad circuit board, etc.

    • tipoo
    • 9 years ago

    The Backblaze guys are on reddit, it seems we may be wrong in assuming consumer use would be easier on the drives than this use. This use is tightly controlled for vibration and heat, while consumer use is not.

    • tipoo
    • 9 years ago

    I’m sure while the concept of mechanical drives seems simple enough, the manufacturing of them is complex as balls. They may not be microprocessors, but it’s also a smaller company than CPU makers. Not to be an apologist, but I mean, if they could make their drives the most reliable I’m sure they would try.

    • ronch
    • 9 years ago

    I’m a proud owner of a Hitachi drive. Hitachi FTW!

    • tipoo
    • 9 years ago

    Not anymore, since they were bought out. Shame.

    • ronch
    • 9 years ago

    Ironically, IIRC Hitachi also tends to be cheaper than both Seagate and WD.

    • ronch
    • 9 years ago

    Is Seagate clueless about the reliability of their drives and how people regard them? Or are they just too stubborn to give a crap about it?

    • swaaye
    • 9 years ago

    Doesn’t surprise me really. I do buy Seagate 7200 RPM drives though because they are so cheap. Have backups and use that warranty if needed.

    • puppetworx
    • 9 years ago

    Seagate coming in last matches my experience in recent years though I’ve been mostly handling 500GB-1TB consumer drives.

    I wish more companies could publish data like this.

    • MadManOriginal
    • 9 years ago

    I didn’t click on the details, but this needs model numbers to be meaningful.

    • Milo Burke
    • 9 years ago

    I’m so glad to read this. Your sample size of one drive proves that no Seagate drives anywhere will ever fail.

    • NeoForever
    • 9 years ago

    LOL you completely miss the point.

    • Rectal Prolapse
    • 9 years ago

    Dunno about the Hitachis, but I have 10 of the Toshiba 3TB variants. I think they’re made in … China? 8 of them still have the Hitachi moniker in their serial numbers.

    EDIT: Actually 8 out of 10 of my Toshibas are reported as Hitachis by my HBAs.

    • l33t-g4m3r
    • 9 years ago

    Who’s really cutting the corners here, Seagate or Backblaze? I think if you actually read the article, you wouldn’t have made that statement.

    • jihadjoe
    • 9 years ago

    Their old drives were actually quite good. My first HDD ever was a 20MB Seagate, and their old 9.8GB Medalist Pro was one of the best drives I had.

    Sad to say they do seem to have dropped the ball with their current lineup.

    • tipoo
    • 9 years ago

    Would you say…I impressed Krogoth?

    • bjm
    • 9 years ago

    GTFO!

    • Krogoth
    • 9 years ago

    Very interesting point.

    • derFunkenstein
    • 9 years ago

    The good news is that I’m backing up everything already, and the really important stuff is backed up in two locations, but man it’s going to be a pain in the rear when a drive dies.

    • tipoo
    • 9 years ago

    I wonder how much of this can be traced back to the manufacturing country of origin…Don’t Hitachis tend to be made in Japan these days?

    • Krogoth
    • 9 years ago

    Impressed by graphs.

    Not really surprised at all. Seagate is more aggressive at cutting corners and can afford to it since they have the largest marketshare in the HDD world. It is like Toyota now and GM back in their heyday.

    • l33t-g4m3r
    • 9 years ago

    [quote<]Only 10% of the hard drives in Backblaze's storage pods come from WD[/quote<] [quote<]Backblaze's purchasing decisions are largely driven by price[/quote<] [quote<]The company uses consumer-grade drives that aren't designed explicitly for RAID environments[/quote<] So what exactly does this study actually show? Don't use consumer-grade drives in RAID environments. That's ALL it shows, and IMO Backblaze made a conscious choice to do so including the possibility of failure, because it was considered cheaper to replace failed drives than purchase RAID specific drives for reliability. FWIW, Seagate's RAID drives on newegg have better reviews than WD, as apparently WD was having failure problems with one of their models. The Barracuda 1.5TB has historically been cheaper than any of the competition while giving better performance, and the reliability is probably average for a 7200 rpm consumer drive when used in an enterprise environment, aside from the initial firmware issues that stirred up the trolls. A lot of businesses do this to save costs, including mine, and IMO it's just stupid. Pay more for the RAID drive and be done with it.

    • hoboGeek
    • 9 years ago

    Yes, exactly. Must be an accident.
    I remember I bought two of those IBM hard drives, one failed after only a month, and the other failed after 13 months. The second one, when I asked for a replacement, I got a Hitachi instead. That last one I still have, while the first one that I replaced with another IBM died right after the warranty (2yrs) ended.
    In fact, all the IBM drives that my friends and/or clients bought have died sooner or later.
    The only striking detail that I remember about those drives was that they were made in Hungary.

    • maxxcool
    • 9 years ago

    WD green == poo. never liked’em … now even less.

    • meerkt
    • 9 years ago

    Perhaps.

    • SoM
    • 9 years ago

    proud Seagate owner, never failed me, currently running 1.5TB @ 3+ yrs straight

    • meerkt
    • 9 years ago

    I think that’s unlikely.

    • meerkt
    • 9 years ago

    It was.

    My replacement 45GB Deathstar drive, after the first one started failing, started collecting bad sectors pretty quickly.

    Though, I did use it for quite a while in a downloading machine, with a procedure to verify downloads and redownload bad blocks as needed. 🙂 It was a bit tedious though, as I never managed to trigger the sector remap mechanism by writing over just the bad sectors: I made a small utility that did unbuffered sector-sized reads. Each sector that took more than N milliseconds was deemed bad, and was overwritten. That didn’t work. I tried reading backwards, thinking there may be some readahead caching, but it didn’t help either. Maybe I should’ve erased a whole area around the bad sectors…

    • meerkt
    • 9 years ago
    • bjm
    • 9 years ago

    Toshiba does now. Western Digital had them initially, but they were forced to sell Hitachi’s assets to Toshiba in order to [url=http://www.ftc.gov/news-events/press-releases/2012/03/ftc-action-preserves-competition-market-desktop-hard-disk<]satisfy the FTC[/url<]. Edit: Format fail. Edit 2: Argh! Edit 3: I hate typing on smart phones. Edit 4: I give up.

    • adisor19
    • 9 years ago

    Disturbingly enough, i still have a Deathstar chugging along just fine as my FreeNAS OS drive.. Must be a freak accident of some sorts.

    Adi

    • Neutronbeam
    • 9 years ago

    All my non-SSD drives have been WD; just never trusted Seagate after anecdotal reviews online. The Hitachi news is very interesting. Good story guys!

    • anotherengineer
    • 9 years ago

    I wonder if any of the Hitachi drive were made in Japan, instead of ‘other’ places? I have seen some laptop drives by Hitachi that were made in Japan.

    • stdRaichu
    • 9 years ago

    Agreed, although I was under the impression that “fact” was a perfectly valid defence against claims of libel (at least in the US). Any IAAL types have any opinions on the issue? It just seems slightly odd to release the study without this set of data and then to add it a couple of months down the line. Anyway, it’ll be interesting to see if and how the HD manufacturers respond.

    The gist I perceived from the google study was that they didn’t publish the names more of a courtesy to the manufacturers than anything else.

    I wonder what the drive manufacturers made of [url=http://infolab.stanford.edu/pub/voy/museum/pictures/display/0-4-Google.htm<]Google's first storage unit[/url<]?

    • shaurz
    • 9 years ago

    I always suspected Seagates were bad so always avoided them. I think they caught the Maxtor lurgie.

    • cmrcmk
    • 9 years ago

    But Western Digital owns Hitachi Global Storage, right?

    • someuid
    • 9 years ago

    IKR?

    (I know, right?)

    • superjawes
    • 9 years ago

    Yeah, I’m going to back up my drive when I get home…

    • superjawes
    • 9 years ago

    It might have more to do with legal reasons, since this could easily affect one’s business. It’s possible that Backblaze only just got rear-end protection for them to release.

    • Omniman
    • 9 years ago

    I’m surprised Hitachi is a high ranking here. I’ve had so many failures per week from them compared to Seagate and Western Digital.

    • derFunkenstein
    • 9 years ago

    Well those Seagate results are going to make me turn a nervous eye to my system. I don’t have the capacity of any of those drives but both of my 1TB drives are from the same family as the 3TB and 4TB Seagates (mine are both ST1000DM003).

    And people think SSDs are unreliable…

    • willg
    • 9 years ago

    This isn’t a typical environment for a consumer drive, so how much of this could be environmental vs. manufacturing or design?

    Could it not be so that the Hitachi drives are more vibration/heat tolerant, or inversely the Seagate drives are more vibration sensitive?

    • Sahrin
    • 9 years ago

    Somewhat ironically, Hitachi is the current owner IBM’s old HDD division, makers of the legendary “Death Star” drives.

    • stdRaichu
    • 9 years ago

    Well this should certainly put the cat amongst the pigeons. I’m curious as to what prompted the change in policy from “no names” to “names and shames”.

    Anecdotally I’d have to agree with the findings of the study, I’ve always found hitachi drives the most reliable with WD a fair-to-middlin’ second. Incidentally I’ve been running a bunch of the WD greens in RAID for aeons and with the newer models I found the spurious failures/drop-outs stopped immediately once you used wdidle3.exe to turn off aggressive head parking.

    • crabjokeman
    • 9 years ago

    Why did you make me look that up on urban dictionary?

    • My Johnson
    • 9 years ago

    IKR?

    • crabjokeman
    • 9 years ago

    [quote<]Infant [b<]morality[/b<] seems to be a bigger problem for the WD drives[/quote<] That's a very unfortunate typo...

Pin It on Pinterest

Share This

Share this post with your friends!